Starting with the Industrial Revolution (or perhaps even earlier), humanity began an unintentional manipulation of the climate of planet Earth, turning up a thermostat we didn’t realize was within our reach. The carbon dioxide pumped into the atmosphere by burning fossil fuels has raised global temperatures, resulting in a host of harmful impacts.

The obvious solution is to stop turning up the heat by eliminating our greenhouse gas emissions. Though that is easier said than done.

But what if we could also take steps to intentionally turn the thermostat down? This is the realm of geoengineering—efforts to intelligently manage the earth’s complex climate system.

There are a number of ways that this could be done, but the one that seems to generate the strongest reactions is the prospect of shading the planet. Since greenhouse gases work by holding in heat energy leaving the planet, you can counteract them by reducing the amount of solar energy coming in.

Such an action would take inspiration not from science fiction (launching giant mirrors into orbit, perhaps), but from volcanoes. Large eruptions can have a cooling effect for several years, because volcanic sulfur dioxide gas forms microscopic particles high in the atmosphere that reflect some sunlight back into space before it can warm the earth’s surface.

If we injected our own tiny particles high into the air to reflect just a little sunlight, we could have an adjustable shade that would counter some portion of our greenhouse warming.

One catch — and there are several — is that the effect is temporary. The tiny particles (called aerosols) will wash out of the atmosphere within a year or so and would have to be constantly replenished.

At first blush, this sort of geoengineering scheme may make you uneasy. Perhaps it just feels “unnatural” and risky.

Perhaps you worry that international conflict or disagreement could disrupt the aerosol replenishment, rapidly bringing back the offset warming. Or, like some critics, maybe your concern is that humanity won’t solve the core problem—greenhouse gas emissions—if a comforting bandage exists.

To the extent that the prospect of solar geoengineering has been discussed, it has been treated as one potential tool in the toolkit rather than a solution to the problem of climate change. It wouldn’t cure the disease, but it might be able to ease some of the symptoms.

Whether it would be worth the risks isn’t clear yet, and climate scientists have differing personal opinions about the idea. But many feel it’s still worth studying, just in case.

Although nations have agreed to a goal of limiting global warming to no more than two degrees Celsius, actual plans to cut greenhouse gas emissions have so far fallen significantly short (even before the United States announced its intention to withdraw from the Paris Agreement). The persistent mismatch between talk and action on emissions has researchers taking an increasingly serious look at methods to hack climate change through geoengineering.

No Free Lunch

Some of the research on solar geoengineering (using computer climate models) has centered around exploring the potential for side effects. Global temperatures aren’t the only thing that have changed due to human-caused warming. So have patterns of precipitation and other weather.

The same would be true of any effort to bring temperatures back down. In particular, shading the planet would be expected to reduce precipitation by a little bit more than warming increases it.

This is because aerosol cooling isn’t an exact match for greenhouse warming. Aerosols help cool the earth’s surface, where solar energy is absorbed, during the day. But greenhouse gases warm at night, too, and exert their influence far up into the atmosphere.

These complications mean that one consideration when designing an actual geoengineering plan would be to evaluate the regional pattern of side effects, which could be stronger or more significant in certain areas. Of course, global warming also has complex regional impacts, so the actual change experienced in any place would be the combined effects of greenhouse warming and aerosol cooling. A little aerosol cooling might only lessen the changes wrought by warming, but there could be other, difficult-to-predict interactions that affect regional weather patterns.

The leading candidate for making aerosols has been the same sulfur compounds emitted by volcanoes. Though such sulfur compounds could come with an additional side effect: reactions with atmospheric ozone that would slow the recovery of the “hole” in the ozone layer.

But recent research has also examined fine calcium carbonate dust as an alternative to sulfur. This technique could cool just as effectively, but with a beneficial effect on the ozone layer.

Potential side effects aside, there are several big-picture things we know about this sort of geoengineering. It could counteract warming, which means it could also help limit sea-level rise by slowing both the melting of glacial ice and the expansion of warming seawater.

But solar geoengineering wouldn’t do anything about ocean acidification — the falling pH of seawater caused by carbon dioxide absorbed from our emissions — and the harm this causes marine ecosystems. That’s one more reason why we could not simply continue to emit greenhouse gases while canceling out their warming with an ever-thicker veil of aerosols.

So, what might a plausible scenario of solar geoengineering actually look like?

Researchers are still working to nail down the basic processes rather than formulating full-fledged schemes. But it seems likely that solar geoengineering would only ever be used with a light touch, if it’s used at all. Alongside any other efforts to halt the growth of greenhouse gas concentrations, a small amount of solar geoengineering could be slowly ramped up (and then down, decades later) to “shave the peak” off global temperatures, for example.

To accomplish this, airplanes might regularly inject aerosols into the upper atmosphere. Because the injections would be phased in gradually, the project could safely be aborted or adjusted if the early results turned up any surprises.

At its strongest, a peak-shaving plan might only knock a couple tenths of a degree Celsius (a few tenths of a degree Fahrenheit) off global temperatures, the equivalent of a decade or two of warming at our current rate. But slowing the warming and reducing the peak this way would help limit the impacts on ecosystems and human systems.

Head In the Clouds

Injecting reflective aerosols into the atmosphere isn’t the only way we could engineer a slightly cooler planet. We could seed low, bright-white clouds, making them even more reflective — a variation on the theme of shading the planet. Or, for a different tack, we could make thin, wispy cirrus clouds higher in the atmosphere even thinner.

That might sound counterproductive, but the water vapor that makes up clouds is a potent greenhouse gas. While low, fluffy clouds can have a cooling influence, high cirrus clouds don’t reflect much sunlight. The amount of outgoing infrared radiation they absorb (which is what greenhouse gases do) is greater than the incoming sunlight they reflect.

Dispersing these cirrus clouds would weaken their greenhouse effect — directly rather than indirectly countering the effect of our greenhouse gas emissions. The technique would be tricky and hasn’t yet been studied as closely as aerosol injection, but it adds another option. It could even be used to supplement aerosol injection and minimize its side effects.

The next step for research on any of these techniques is to move from virtual experiments to physical ones. Small-scale pilot tests will be needed to understand exactly how things would work. Only then could we get closer to actually engineering “treatment plans” that world leaders could seriously consider.

A Harvard research group is planning a small pilot experiment soon that would involve releasing aerosol particles from a high-altitude balloon and analyzing their behavior. Plans for a similar experiment in the UK were canceled in 2012, so this would be the first outdoor test of the technique. (Although an ongoing University of Washington project studying cloud brightening also aims to include experiments.)

David Keith, the Gordon McKay professor of applied physics and professor of public policy at Harvard, describes these tests with the language of due diligence. “There is a large body of evidence now from pretty much every climate model we have, and basic science, and also high-level consensus agreements like the U.S. National Academy and the UK Royal Society, that these technologies might be able to significantly reduce climate risks, albeit with lots of uncertainty and potential new risks,” says Keith. “The main argument is simply that we should get serious about knowing more — that ignorance is a bad thing [and] dangerous on topics like this. It could be that what we’ll find out is that it’s a bad idea, and we shouldn’t do it, but that’d be really useful.”

The effort to stabilize the earth’s climate can be visualized with individual wedges that stack to push future temperatures below the current warming trend. Each wedge is a contribution from a different factor, like improved energy efficiency in industry, electric vehicles, or the transition to clean energy. It’s possible that solar geoengineering could put another wedge in humanity’s back pocket.

But whether we decide to actually try it out, while we seek to undo our unintentional geoengineering experiment of burning planet-warming fossil fuels, remains a very big question mark.